Long GOP VS All-I (Can YOU Tell The Difference?)
ฝัง
- เผยแพร่เมื่อ 23 พ.ค. 2023
- All Example Footage Graded With Phantom LUTs:
transactions.sendowl.com/stor...
THE GEAR I USE:
*Some of the following links are affiliate links, so if you decide to purchase something, I get a very small commission at NO extra cost to you - Thanks in advance for supporting the channel!
CAMERAS:
My Main Cam (Lumix S5IIX)
geni.us/VHW0Kv
My B-Cam (Lumix S5II)
geni.us/Avt5
PRIME LENSES:
Panasonic LUMIX 24mm F1.8
geni.us/AdbkotR
Panasonic LUMIX 35mm F1.8
geni.us/j0l5fn
Panasonic LUMIX 85mm F1.8
geni.us/mZm1TkS
ZOOM LENSES:
Sigma 28-70 F2.8
geni.us/Bd6NL
Panasonic LUMIX 16-35 F4
geni.us/oTkQR
Panasonic LUMIX 70-200 F4
geni.us/RBByZLM
AUDIO SETUP:
Rode NTG4+
geni.us/2jHAaA
Tascam DR60 MKii
geni.us/fXkZU5S
Hollyland Lark 150
geni.us/6tla19
The Lumix S5IIX is the first camera in the S5 lineup to have ALL-Intra codecs internally.
This makes the Lumix S5IIX a solid choice for most professional videographers... However, can you really tell much of a difference between All-I and Long GOP?
In this video, I delve deeper into the Long GOP and ALL-I codecs that are found in the Lumix S5IIX, and there's some image quality comparisons along with a post-workflow stress test as well.
If you're unsure about whether to buy the Panasonic Lumix S5II or Lumix S5IIX , then this video could help! - เพลง
Could you guys tell the difference between Long GOP and All-Intra?
I personally wasn’t!!! It could be the TH-cam compression too!!!
Yeah.. some of the places with high movement seemed a bit mushier in Long GOP. With my GH5 it was definitely noticeable so I would use All-I for fiction projects and Long GOP for documentary work. Difference is minimal, though, so you can definitely get by using Long GOP for most regular work
Long gop is heavily compressed all i is much more easy on your computer also
I wish I could spot the difference 😂I guess the TH-cam compression doesn't help, either that or my eyes aren't as good as I thought
My understanding is, that this is about the original footage coming from the camera. When it's processed, it is Long GOP anyway. Re-watching it from TH-cam is even more destroying any differences.
I also think that a lot of movement typically is blurred by 180° shutter angle. So there isn't much to see where All-Intra shines other than in the advantage while processing the footage.
My expectation is that: if you have a detailed background that is in focus and you have an object in front that is out of focus that moves out of the frame, All Intra might show visible artifacts for a few frames in the areas where the background comes into sight. (A train leaving the station, revealing the scene with people behind the train).
It might be interesting to see if this is clearly visible in the original footage and if the difference survives the TH-cam-compression.
All-Intra might be a huge advantage if you use very short exposure in action sequences like sports and you want both - video and extract frames as stills.
been wanting to see all-i vs long gop comparison on the s5m2x for a while and my man came through with the perfect video
Keepin it real since 2020
I think you'll still be waiting as TH-cam compresses the video so you cant see a difference.
@@kevinforget549 so you do see a difference on monitors then? How big of a difference would you say it is?
literally satisfied my demands.
@@zhongjiezheng4646 you should as it isnt compressed as much. Ive never actually used all-I before and my s52x is still on backorder but ill definitely be testing it out, especially the 6k raw from the Atomos recorder which is the same as 4:4:4:4 vs the 6k 4:2:0 from the internal.
Excellent video! I'm planning on doing a deep-dive codec analysis video when my S5iiX comes in, comparing BRAW, ProRes RAW, internal and USB SSD codecs. Plus quantifying noise levels.
Great video Josh! Really brilliant & clear explanations.
Thanks Josh. Excellent, like usual
Thanks Patrick, I appreciate it!
Where you can see an actual difference is with external BRAW.
Fixes gamut clipping, green shadows and banding issues and almost one extra stop on top with highlight recovery enabled.
Thanks for the info
Thanks for answering this since the creator of the video didn't really show images for comparison purposes
This was great. I'd love to see a video comparing 4:2:0 10 bit vs 4:2:2 and raw (4:4:4 theoretically) in specific color resolution tests to see if there is any situation where you can actually see a difference. I suspect that the only real difference that's detectable is 8 bit vs 10 bit vs 12 bit. And chroma subsampling never is noticable, even after editing, since delivery on TH-cam is 4:2:0.
I'd love to make that video for you! It's always frustrating when I make videos comparing codecs etc for TH-cam because it looks great on my 4K monitor, and then once it hits TH-cam all of those subtleties disappear !
You have some of the best tech reviews out there but I would enjoy seeing a short film or vlog from you using the S5ii and Lumix primes. You make epic work. Cheers.
Appreciate this so much , thanks man! Will defo make something of that sort in the near future :)
Super helpful! Thank you!
Amazing. Thank you!
Hi Josh ...Thank you for producing great videos on the Panasonic Lumix S5-iiX ... CODEC's and cage build for starters... .....it's a fantastic camera, which I am using on a feature, that I'm DOP on .... keep it up 👍👍👍
Thank you very much! Hope you manage to make the most of this incredible camera :)
You so clever. That was really well explained thanks Josh!
Thanks for making this!
My pleasure :)
All intra looks so much better!! I’ve been waiting for this!! Great work man, thx a ton
Second 🎉
Love your videos about Lumix cameras always very informative and helpful content 👍🏻
Please make a video about ‚the perfect settings for video‘ for the s5iix
Keep up the good work
Greeting from Germany
Thanks for the kind words, really appreciate it! I'll put that video idea on the content list. Thanks and all the best!
Great video as someone that just moved to Panasonic and got the S5ii and have no choice but to use long gop & have a MacBook M1 Pro chip also so this video was perfect not sure if I will upgrade to the S5iix I have to use my S5ii for a few months first since I’ve used the canon eos r & used ALL-I mostly this is good information to know thanks for making this comparison video
Thanks for the vid! It is an interesting debate on price do you spend more money on the better laptop and save all the $$ you would spend on SSD’s? I’ve had no problem at all with Long GOP on the M1 Pro it handles it so easily and my clients have been happy to save the hard drive space bc they back everything up on the cloud and THAT gets even more expensive than hard drives.
The best test Josh is trees and forests. Lots n lots of detail. Or use a death chart to check the PQN between codecs.
I've had a unique case use in shooting roller coasters to be slowed down in post. All-I is superior but fps becomes the limiter at around 50mph/80kph. Beyond that speed, motion clarity starts to break down and I've had to use S&Q at 120 or 150fps (10bit/All-I/422/30fps).
Thanks for your Video,
Yeah it is more a question if you need S5iiX or S5ii is enough...nice choice of music by the way
Yours Fil big fan of your channel😊
Thanks! Hahaha yeah I love the 80's kinda synth vibes!
That USB-C cable all crooked on the side there almost killed me :) good video. I didn’t know that was how GOP behaved :)
hahah same here ...that was the first thing that caught my eye
Hahaha since this video i've got a new cable that works a bit better :)
@@JoshCameron okay, great ..now I can sleep again😛
Making me wanna buy the lumix s5 sharpie edition 🔥
Hi, tks for the video. Can you make a video testing the steaming (wired and wireless) capability of the sii X. The streaming capacities are the only one that would make me choose the x over the s5ii. I better be sure there working OK because so far the reviews are so so on this point. Tks
Good video. Thanks. toronto canada.
Really helpful, thanks! How about 6k Long GOP footage though on that m1 machine? Would love to get my hands on some sample footage to download and test myself.
Works beautifully mate, you have nothing to worry about there. The M1 machines handle every codec from these cameras very smoothly :)
There is no way to change the GOP parameters in these cameras, but is there at least some information about how good the video encoder is in detecting scene changes. Recording only differences when the entire image changes will not be efficient, so the video encoder normally inserts a new I-frame when this is the case, it would be interesting to know how the Panasonic video-encoder is programmed in that regard, and how long the GOP-sequences are by default.
All-I definitely looks better but is it worth it for people on tight budgets when i comes to storage?
I would've liked if you explained and pointed out a little more of these artifacts. I was barely able to see them and confused if I was focussing on the right thing especially considering youtube compression. Maybe zoom in and show where the differences are?
Also the conclusion about editing on an older machine that you might need upgrading is missing a part where you'd consider a proxy workflow. I'm not sure if FCPX even has that option but for most people proxies would be the more economic option
how do i convert long gop codec to display my video in davinci not only showing an audio cause when i record on hlg on lumix gh5 it dont show in davici
visually there is zero difference to the naked eye, especially on TH-cam.
Great video Josh!
That totally depends on what you view this with. On a 85 inch 4k tv there most certainly is a difference, the long gop looks mushier at times.
@@cmxpiipl absolutely! from the original files but not from TH-cam. That's not only long-GOP but also heavily compressed.
Also how many people have 85" 4k TV?
@@PavSZ I meant I can see the difference from the youtube video. But you are right that an 85 inch screen ain't the norm but a proper stress test
Hey Josh! Great video as always. Would you please make a video of the s5iix with blackmagic raw ? Thanks
It's on my to do list! Thanks dude :)
@@JoshCameron you’re the best!!
I spoke with Panasonic's codec developer a couple years ago, he was 100% confident that there is no difference in the quality of shots, it's merely a matter of saving space vs ease of editing. Unless it's uploaded on tv, then Long gop has more issues
Source: Trust me bro
I would be more interested in seeing the difference when keying green screen for example.
All-i is sharper and cleaner read out. Colors look better to my eye as well. S5 iix for me.
Nice one dude - the S5IIX is a great camera, in fact it's the best camera I've ever used :)
2:15 Do You see a difference before youtube compression? and how big is it?
On my GH6, when I turn on the display - WFM/Vector scope it disables the Histrogram why?
For some reason I'm seeing some artefacts on all S5II and S5IIX videos including my own S5IIX when recording internally. I come from a Canon C70 and I did not have this issue or even in sony cameras. However, for what I see it does not happens in all videos recorded in BRAW. Does anyone have tested this?
Honestly I’d just use BRaw since I’d be using an SSD anyway.
I do wish I got the S5IIX though but couldn’t wait for months since my original S5 would depreciate and not get me enough money for the higher end one.
Not sure I understand what the conclusion is? Of course it's a tradeoff between quality and size/processing time. But, can you tell the difference or not?
Hi how much memory size is for 5min video of all intra 24fps 4k 10bit 4:2:2?
I shoot weddings on my Pansonic GH6 should I use Long GOP or All-I
Would also LOVE to see a video on RAW vs Long GOP V-Log!
RAW is way superior. Trying out BRAW at the moment and it almost looks like a completely new camera.
@@DennisSchmitz Always superior RAW but just wanted to know how drastic of a difference. What do you like about BRAW more than VLOG RAW?
@@joeensley3912 quite drastic tbh.
No green tint in darker areas
No banding in shadows
No abrupt shadow crushing
No gamut clipping with super saturated reds and purples
Better texture and detail
No temporal NR slow motion looking noise patterns
@@DennisSchmitz Better color science too (I know people love to dismiss the importance of a good starting point). Ive noticed since my S5, that the Prores raw isnt as good as BRAW, because prores raw keeps the same color science while BRAW adds their usual flare to it. The skin tones and cyan blues ive grown to absolutely love.
I compared them at 24fps on my S1H and didn't notice any difference. Im sure theres a difference but i didn't notice them in my samples.
How old was the old mac?
After watching your video I switched all the format to LONG-GOP, The minor differences in quality (if they even exist) is not enough to compensate the 3X file size differences.
The only issue I have with this video is that on the older machine rendering was completed where on the M1 machine it wasn’t.
I don’t think long GOP would have scrubbed that easily on the Intel machine had it not finished rendering.
You are showing the nitrate with MB instead of Mb or mb. There is a huge difference between those mate.
Cheers.
I have a little hope they will bring all-I to the s5ii version. If not I’ll have to figure out how to sell the s5ii for a decent price and get the x. Everyone says they won’t, but I don’t see why they wouldn’t
There is an update for 200$ for the s5ii and i think it includes all intra
@@marionxah that’s would be sick if it was internal! I think it’s for raw to external :(. Correct me if I’m wrong
@@hhankx idk haha but i am thinking of selling my S1 and get a S5iiX
Нужно было также приложить пример минимального битрейта для 4к разрешения, которого для большинства случаев более чем достаточно
Any difference in low-light? Might be worth it if it's less noisy.
Same noise performance I'd say !
I was gonna say the example footage looks so much better than your usual work, those phantom luts are actually pretty good then
I have seen bigger differences between a good and a bad narrative
long gop is the way to go. The file size trade off for the performance loss (I mean is there really any loss), is more than worth it. In my opinion, if you’re shooting all intra for “performance” reasons, you’re mis-spending your money. I mean who the heck has a camera that can do crispy 4K 60 10 bit in all intra but their computer can’t play it for editing, AND it has tons of extra storage to manage those more than double sized files?
All intra is only for professionals who are explicitly asked to deliver all intra contractually.
Yeah I get what you're saying and agree with you. All Intra is great and I do use it quite a bit on my client shoots now, but Long GOP is incredible as well, and the smaller file sizes make it easier to manage :)
I use All-I everytimes on my GH5, GH6 but I don't mind using Long GOP on my S1 and S5II, the differences are most of the time neglibile.
However the detail rendering is much more important for me (no noise reduction nor oversharpening). In this area, the S5II is a big disappointement with the standard profiles (rec 709).
Hi Josh, the speed units are incorrect, they should be in Mb(it)/s and not MB(yte)/s (th-cam.com/video/Ve-99RLmbY0/w-d-xo.html) - I have the exact older Macbook Pro like your are comparing to in this video and if you go to an intermediate workflow or with proxies, the old Macbook just works fine. Nevertheless I still haven't found out how to tell if a h.264/h.265 encoded MP4 file is intra or interframe encoded, do you maybe know? Mediainfo will not show any indication...
does this camera shoot in 24fps?
If you switch to the correct region for it :)
Pleasse make a version comparing this with proress
It's coming!
And slow motion zoomed in at 400 percent on a detailed fast moving image
Great stuff!!! Great explanation with PRACTICAL implications! However, at 3:29 you hypothesize a reason for slower editing, between GOP and Intra-I, based on file size... Why? If you follow your OWN logic about how each compression approach works it's obvious that in Long GOP the computer has to go backward and forward to assemble each frame. That's what slows the computer down. In a sense, for each frame of image it works on it need to manipulate the data of 30 images, say. If you scrub ahead one image, it may have to go back 30 frames to construct it.
In Intra-I, as you point out, each frame is compressed on its own. There's no need to go back and forth through other frames. Also, and I speculate here, JPG compression is built into the hardware, in a sense, of most computers these days--and they optimize the __ out of it. And I believe Intra-I is basically that. So your speed benefits from that too.
Again, great video!
The skin tones are more natural in All-i than the long-gop shot.
What's the point of having 4k 60fps in crop 😢?
Whats the problem? Just use a wider lense if you dont like it.
I think it's diminishing returns at this stage. I think a few years ago ALL-i was a big deal but now computers are better at handling the Long Gop. You don't even need to top spec, just half decent will do. But this all depends on how intensive your editing is. Are you a professional editor doing very intensive work with many layers and effects or do you do the odd video for fun. I think if you're a pro get a pro machine to edit on AND get the ALL-i to make your work flow as fluid as possible. If not update your machine as it will probably be doing more than just video editing and more power is never a bad thing.
looks like all i is a tiny bit cleanerin terms of image
In some cases definitely :)
real world difference is the storage you need to pay for shooting flowers and grass in he mothers backyard. hahaha.
anyway, i think green screen footage would benefit a lot from all-i.
GOP = Group of Pictures (not photos)
Same difference :)
It must be possible to record blackmagic Raw to SSD, the problem is the patents of the stupid Red company.
If they were my neighbour i would know how to thank them, perhaps by sending them a rusian cocktail.
Don’t forget to make it extra spicy
Looks like Nikon vs Red is going to help with this. I would love BRAW internal, just the best codec I have used easily.
@@jessecumberledge4530 lets make panasonic sue red! lets ask NIKON for the sauce of their victory! FOR RAWWWWWWWWWWWWWWWWWWWWWW
I must disappoint you, my friend. PCs have always been faster than the Macs.
I'm fairly certain my 60-pound PC does the job faster than your cute laptop ;)
Ps.
My PC plays perfectly smooth 4k videos in real-time as well as it converts them to H264 in around 10-15 sec per 1 min of 4k footage, so we PC users are good :D
I must disappoint you, nobody cares 😅
You certainly do care enough to reply to my post, or was it just a small ego trip? I thought so 💓 @@ValerioCata
Sorry but pc users are probably more than mac users. In my country an m1 Mac pro laptop (2021) costs around 3,500 euros today. With 4000 euros you can have a pc, with i9 13900k + rtx 4090 + 64gb. What I want to say is that only from a price of 3,500 and above, you can say that you can use log gop files !!! But with all-intra or proress, YOU CAN ALSO WITH AN OLD pc (5 years old). Bottom line if you want panasonic the s5iix IS ONE WAY.
Sorry dude but why you speak at Doppel time? Information should be understood from the brain. I really wish I could keep looking because you're a likeable guy. but sorry i'm out
Hey dude, I’ve had a lot of people in the past say that I talk quite quickly in my videos and it’s something I’m actively trying to get better with; but I actually speak at this speed in my every day life too 🤣😂 so I think it’s something I have to train within myself unfortunately. I appreciate you leaving this comment and I apologise for the speed of my speech, but please know that I’m working on it 👍🏼
That means no difference cause both quality have lowest bitrate from exporting default fcpx setting.
Higher bitrate = higher resolution quality
Of course nobody know this, use custom output settings from comressor